Introduction
About 85% of thyroid cancer cases are papillary thyroid cancers (PTCs), which are predisposed to metastasize to lymph nodes in the neck, increasing the likelihood of recurrence and adversely affecting prognosis [
1]. The metastasis of lymph nodes in the neck is a critical indicator of PTC progression, directly determining the surgical approach and the formulation of a postoperative comprehensive treatment plan [
2]. Therefore, examining the internal driving forces of cervical lymph node metastasis in PTC is fundamental to finding a suitable solution.
circRNA is a non-coding RNA that forms a loop structure by covalently closing the 5´ and 3´ ends of the precursor RNA, most of which is localized in the cytoplasm, and its highly conserved and non-easily degradable structure by RNA exonucleases makes its expression stable [
3]. The dysregulation of circRNA affects cell proliferation, apoptosis, epithelial–mesenchymal transition (EMT), invasion, tumor immune evasion, gene mutation, and modification, ultimately promoting tumor progression and invasive metastasis, thus becoming a potential biological marker for tumors [
4].
Tyrosine kinase II (CSNK2) is a serine/threonine kinase that phosphorylates hundreds of substrates. CSNK2 consists of two catalytic subunits (CK2α and CK2α´) encoded by CSNK2A1 and CSNK2A2, respectively [
5], and two regulatory subunits. Reportedly, CSNK2A1 is expressed at abnormally high levels and kinase activity in a variety of cancer cells [
6‐
10], and it stimulates tumor proliferation, DNA damage repair, EMT, drug resistance, and other biological behaviors by phosphorylating key molecules in various signaling pathways [
11]. CSNK2A1 has also been found to regulate Akt activity by phosphorylating Akt, leading to the hyperactivation of the PI3K/Akt signaling pathway [
12].
In this study, we used high-throughput sequencing and identified a total of 17 up-regulated circRNAs in extensive lymph node metastatic PTC compared to tissues without extensive lymph node metastasis and then selected circNDST1 as the key subject of our investigation. We established the key role of circNDST1 in enhancing the proliferation, migration, and invasion of PTC cell lines in vitro and in vivo. Notably, we provided evidence that circNDST1 binds to and promotes the function of CSNK2A1. CircNDST1 stimulates Akt signaling pathway activation and EMT by relying on CSNK2A1’s mediation.
Methods
Human papillary thyroid cancer tissue specimens
PTC tissue specimens for this study were provided by patients who had their PTC surgically resected at the Union Hospital of Tongji Medical College. Cancerous tissue specimens were collected from six patients with thyroid cancer. Clinical information on the patients is shown in the attached table. Of the six patients, three had extensive lymph node metastasis, and the other three had no extensive lymph node metastasis. Tissue specimens were immediately stored in liquid nitrogen until use. Written informed consent was obtained from all patients, and the investigation was approved by the Ethics Committee of the Union Hospital of Tongji Medical College.
RNA sequencing of circRNA extracted from human papillary thyroid cancer tissue
RNA was extracted from 3 cases with extensive metastasis in the neck lymph nodes and 3 cases without extensive lymph node metastasis in the thyroid glands. A de-ribose-specific library was then created and sequenced utilizing the Illumina HiSeq2500 platform (Illumina, San Diego, USA) according to the manufacturer’s protocol.
Cell lines and culture conditions
Three PTC cell lines, TPC1, KTC1, and BCPAP, and normal thyroid epithelial cells, Nthy-ori3-1, were purchased from ATCC. TPC1 cells were cultured in a DMEM medium (Gibco, Carlsbad, CA, USA), while KTC1, BCPAP, and Nthy-ori3-1 were cultured in an RPMI1640 medium (Gibco, Carlsbad, CA, USA), both of which contained 10% fetal bovine serum (Gibco, Carlsbad, CA, USA). The cells were maintained in a humidified incubator at 37 °C and 5% CO2.
Fluorescence in situ hybridization (FISH)
Cy3-labeled circNDST1 FISH probes designed by Guangzhou RiboBio were used to determine the localization of circNDST1 in cells. FISH was performed with a fluorescent in situ hybridization kit (RiboBio, China, cat. NO: C10910) according to the manufacturer's protocol. Nuclei were stained with 4,6-diamidino-2-phenylindole (DAPI, Beyotime, China, cat. NO: C1005), and photographs were taken with a fluorescence microscope.
RNase R treatment
RNase R treatment was carried out according to the manufacturer’s protocol. 10 μg of total RNA was incubated with or without 10 U/μl RNase R (BioVision, USA, cat. NO: M1228) for 15 min at 37 °C, treated at 70 °C for 10 min to inactivate the Rnase enzyme, and the contents of circNDST1 and NDST1 in the RNA samples were evaluated using RT-qPCR.
Actinomycin D assay
TPC1 cells were treated with actinomycin D (Selleck, USA, cat. NO: S8964-01) for 0 h, 4 h, 8 h, and 12 h before RNA extraction. Actinomycin D's working concentration was 5 μg/ml. The contents of circNDST1 and NDST1 in the RNA samples were assessed using RT-qPCR.
Cytoplasmic and nuclear RNA fractionation
Nuclear and Cytoplasmic Extraction Reagents (ThermoFisher, CA, USA, cat. NO: AM1921) were used for nuclear–cytoplasm separation before RNA extraction for RT-PCR. GAPDH and U6 were employed as the respective positive controls for cytoplasmic and nuclear RNAs.
Sanger sequencing
The amplification products of circRNA were inserted into a T-vector for Sanger sequencing by Sangon (Shanghai, China). A primer (Tsingke, Nanjing, China) was designed to confirm the back-splice junction of circNDST1.
RNA extraction and real-time fluorescence quantitative PCR
RNA was extracted from tissues or cells using a Trizol reagent (Vazyme, Nanjing, China, cat. NO: R401-01) and reverse transcribed to cDNA per the protocol of the PrimeScript RT Reagent Kit (Vazyme, Nanjing, China, cat. NO: R323-01). Real-time quantitative PCR was used to determine the expression of target molecules. PCR primer sequences were synthesized by Tsingke (Nanjing, China) and are listed below. RNA expression fold changes were determined utilizing the 2−ΔΔCt method.
GAPDH Forward Primer: 5´AGAAGGCTGGGGCTCATTTG 3´
Reverse Primer: 5´AGGGGCCATCCACAGTCTTC 3´
CircNDST1 Forward Primer: 5´ CCGCTCTGGCAGGTTCT 3´
Reverse Primer: 5´ GTTGGACAGGTGCGTCAT 3´
NDST1 Forward Primer: 5´ TTTGTTGGTCAGTGGACGATT 3´
Reverse Primer: 5´ CAGAAGATGAACAGCAGGAAAA 3´
U6 Forward Primer: 5´CTCGCTTCGGCAGCACA 3´
Reverse Primer: 5´AACGCTTCACGAATTTGCGT 3´
18S Forward Primer: 5´GTAACCCGTTGAACCCCATT 3´
Reverse Primer:5´ CCATCCAATCGGTAGTAGCG 3´
CSNK2A1 Forward Primer: 5´GAACGCTTTGTCCACAGTGA 3´
Reverse Primer: 5´TATCGCAGCAGTTTGTCCAG 3´
DNA was extracted from cells following the protocol of the TIANamp Genomic DNA Extraction Kit (cat. NO: DP304-03) and used for subsequent RT-qPCR assays.
Western blot analysis
Cells were washed with PBS, lysed in a RIPA protein extraction lysis buffer (Biosharp) containing protease and phosphatase inhibitor cocktails (MedChemExpress), and their total protein concentrations were determined with a spectrophotometer at 562 nm via the BCA assay (Vazyme, Nanjing, China). The supernatants from cell lysates were run on 10% or 12.5% acrylamide gels using SDS-PAGE and then transferred to NC membranes (Millipore). The membrane contents were blocked with 5% skim milk in Tris-buffered saline and Tween 20 (TBST) at room temperature for 1 h and incubated with antibodies at 4 ℃ overnight. Antibodies against β-actin (1: 1000, Cell signaling Technology, cat. NO: 8457 s), E-cadherin (1:1000, ABclonal, cat. NO: A3044), N-cadherin (1:1000, Proteintech, cat. NO: 22018-1-AP), Vimentin (1:1000, Proteintech, cat. NO: 10366-1-AP), Snail (1:1000, ABclonal, cat. NO: A11794), Akt (1:1000, Cell signaling Technology, cat. NO: 4691 s), p-Akt (Ser473) (1:1000, Cell signaling Technology, cat. NO: 4060 s), phosphatase and tensin homolog (PTEN) (1:1000, Proteintech, cat. NO: 22034-1-AP), p-PTEN (Thr382/383, Proteintech, cat. NO: 29246-1-AP), p70 S6K1 (ABclonal, cat. NO: A2190), p-p70 S6K1 (Thr389, ABclonal, cat. NO: AP0564), CCND1 (1:1000, Cell signaling Technology, cat. NO: 2922 s), CSNK2A1(1:1000, Proteintech, cat. NO: 10992-1-AP), and an HRP-conjugated secondary antibody (1:3000) were employed for western blotting. Proteins were identified with a chemiluminescence western blotting detection system (Bio-Rad). Three Western blot experimental replicates were conducted for each molecular assay.
RNAi and cell transfection
SiRNAs (si-circ#1, si-circ#2, si-circ#3) and siNC (RiboBio, Guangzhou, China) designed for the back-splicing sites of circNDST1 were used to specifically knock down circNDST1 without affecting the linear NDST1 molecular expression. The siRNA sequences were as follows:
si-circ#1 ACCCGCTCTGGCAGGTTCT;
si-circ#2 GCTCTGGCAGGTTCTCCCA;
si-circ#3 TCTGGCAGGTTCTCCCACG.
The siRNA sequences used to knock down CSNK2A1 were as follows:
si-CSNK2A1#1: GTCAGCAGCGCCAATATGA;
si-CSNK2A1#2: GGTGAGGATAGCCAAGGTT;
si-CSNK2A1#3: GTTTGGATATGTGGAGTTT;
Transfections were performed with Liposome 3000 (Invitrogen, Carlsbad, CA, USA, cat. NO: CA92008) according to the manufacturer's protocol. Stably transfected TPC1 cell lines used to knock down circNDST1 were constructed with a lentivirus purchased from Genechem (Shanghai, China) and were screened with puromycin (Invitrogen, Carlsbad, CA, USA, cat. NO: J67236.XF).
EdU incorporation assay
The EdU assay was conducted using a Cell-Light EdU DNA Cell Proliferation Kit (RiboBio, Guangzhou, China, cat. NO: C10310-1). TPC1 cells were incubated with 50 mM EdU for 2 h, fixed in 4% paraformaldehyde, and stained with an Apollo Dye Solution, and their nuclei were identified using Hoechst 33,342. Proliferation-positive cells were photographed and counted under a fluorescence microscope.
CCK8 (Bimake, cat. NO: B34304), colony formation assay, transwell migration and invasion assay, and wound healing assay were performed as previously reported [
13,
14].
Biotin-coupled probe RNA pull-down assay and mass spectrometry
A biotinylated circNDST1 probe (5´ CTGCCGTGGGAGAACCTGCCAGAGCGGGTC 3´) and a control probe (5´ GACCCGCTGCTGGCAGGTTCTCCCACGGCAG 3´) synthesized by Sangon (Shanghai, China) were used to pull down the protein bound to circRNA. Approximately 1 × 107 TPC1 cells were lysed and incubated with biotin-labeled probes, and the biotin-coupled RNA protein complex was pulled down using streptavidin affinity-coated magnetic beads (MedChemExpress, cat. NO: HY-K0208), which were then scrutinized with mass spectrometry (SpecAllly Life Sciences Co. Wuhan, China) to identify the protein bound to circNDST1.
RNA immunoprecipitation (RIP)
The RIP assay was carried out using a Magna RIP RNA Binding Protein Immunoprecipitation Kit (Millipore, USA, cat. NO: 17-700) according to the manufacturer’s instructions.
Transcriptome RNA sequencing
Transfected TPC1 cells were lysed with Trizol, and the changes in the transcript levels of TPC1 cells in the lysate were determined by Haplox (Jiangxi, China) after the knockdown of circNDST1. KEGG pathway aggregation analysis and transcriptome volcano mapping were performed using Sangerbox tools, a free online platform for data analysis (
http://vip.sangerbox.com/).
Protein immunoprecipitation
Approximately 1 × 105 TPC1 cells were lysed with an IP lysis solution (Beyotime, Shanghai, China, cat. NO: P0013) and incubated separately overnight with 5 μg anti-CSNK2A1 antibody and IgG (Proteintech). The antibody-protein complex was then pulled down using protein A/G magnetic beads (MedChenExpress,cat.NO:HY-K0202), and protein molecules bound to the target protein were spotted using western blot.
Nude mice tumorigenesis assay
Four-week-old female BALB/C nude mice were selected for the xenograft experiment and kept under specific pathogen-free conditions. The experiment was approved by the Institutional Animal Care and Use Committee of Huazhong University of Science and Technology (Ethical number: S2841).
Overall, 10 nude mice were used for the xenograft experiments: 5 mice in the experimental group and 5 in the control group. Approximately 1 × 109 TPC1 stably transfected cells mixed with a matrix gel (Corning, USA, cat. NO: 354234) were injected subcutaneously into nude mice. Tumor volumes measurements (length × width2/2) started 2 weeks after inoculation. The nude mice were executed 6 weeks after inoculation with cells, the tumors were removed, and the tumor volume and weight were measured.
Statistical analysis
Statistical analyses were conducted primarily using SPSS 21.0 (IBM, SPSS, Chicago, IL, USA) and GraphPad Prism 6.0 (GraphPad Software Inc., CA, USA). Differences between groups were analyzed using Student’s t-test and one-way ANOVA.
Discussion
The rate of thyroid cancer incidence is increasing year after year, the vast majority of which is PTC [
19,
20]. Despite the good prognosis of PTC, cervical lymph node metastasis remains a determining factor for surgical choice and distant metastasis. Therefore, outlining the underlying molecular mechanisms that lead to cervical lymph node metastasis in PTC is critical to the development of new anti-tumor strategies.
circRNA affects tumor development in a variety of tumors, including breast cancer [
21], colon cancer [
22], renal cell carcinoma [
23], and bladder cancer [
24], through various mechanisms influencing tumor cell proliferation, migration, invasion, DNA damage repair, epigenetic modifications, and drug resistance. In this investigation, three PTCs with extensive neck lymph node metastasis and three PTCs without extensive lymph node metastasis were high-throughput sequenced for key circRNAs involved in neck lymph node metastasis in PTCs. Hsa_circ_0006943 (circNDST1) was identified as the most decisive target. CCK8, plate cloning, EdU, transwell, and scratch assays, as well as tumor formation in nude mice, revealed that circNDST1 promoted thyroid cancer proliferation, migration, and invasion.
Our study provides a rigorous approach to studying the function of circRNAs. CircRNAs are essentially formed by the exon and intron splicing and cyclization of their linear molecules, suggesting that when knocking down circRNAs, siRNAs can easily produce off-target effects and incorrectly knock down linear molecules. Because subsequent meticulous functional studies did not confirm that the anticancer effect is produced by circRNA alone, si-circ#1, which did not upset the linear NDST1 molecule, was selected for functional experiments when knocking down circNDST1. However, we did not expand the clinical sample size to fully validate the correlation between circNDST1 and thyroid cancer lymph node metastasis, and that is a dent in our research.
EMT confers metastatic properties to cancer cells by enhancing tumor cell migration, invasion, and resistance to apoptosis. This process is, therefore, considered a marker of carcinogenesis [
25,
26]. Additionally, the activation of the PI3K–Akt signaling pathway enhances tumor cell invasion and oncogenic gene expression via the phosphorylation of Akt [
27]. PTEN inhibits tumor invasion by suppressing the PI3K–Akt signaling pathway and counteracting its cascade response [
28]. In the present study, we demonstrated that inhibiting circNDST1 expression decreased the phosphorylation levels of Akt, p70 S6K1, and PTEN (a tumor suppressor) and increased the expression of PTEN, thus preventing the activation of the PI3K–Akt pathway. CircNDST1 also affected the expression of E-cadherin, N-cadherin, Vimentin, Snail, and other key molecules of EMT. CircNDST1 is a pro-oncogenic molecule of the PI3K–Akt signaling pathway and EMT.
CSNK2A1 has been shown to affect tumor recurrence, metastasis, and prognosis through the phosphorylation of various substrates, making it a key target for anti-tumor therapy [
6,
29‐
32]. CSNK2A1 encodes the catalytic subunit of protein kinase CK2 and enhances Akt activity by phosphorylating Akt. It also phosphorylates PTEN, which reduces PTEN stability and alters PTEN localization in cells, thereby activating the PI3K–Akt pathway to promote tumor cell growth, adhesion, and migration [
5,
11]. In the present study, we queried the binding of circNDST1 to CSNK2A1 via biotin-coupled probe pull-down and RIP assay and also confirmed CSNK2A1’s pro-carcinogenic role in thyroid cancer through a series of functional experiments. Immunoprecipitation and western blot revealed that CSNK2A1 activates the PI3K–Akt pathway by interacting with Akt and phosphorylating Akt. Our subsequent speculation that circNDST1 activates the PI3K–Akt pathway by reinforcing the binding of CSNK2A1 to Akt was validated by IP experiments after the knockdown of circNDST1. CSNK2A1, a kinase, becomes a hub between the circNDST1 and PI3K–Akt pathways. This mechanism also gives rise to new concepts for circRNA mechanistic studies, bringing CSNK2A1 to the fore as a new target for thyroid cancer.
In the present study, we demonstrated that has_circ_0006943 promotes the development of thyroid cancer by binding to CSNK2A1, facilitating the binding of CSNK2A1 to Akt, and subsequently activating PI3A–Akt and EMT. To sum up, by using circRNA high-throughput sequencing and functional validation, our investigation has demonstrated the potential of circNDST1 as a prognostic biomarker for thyroid cancer. In particular, the cyclization properties of circNDST1 and its stability offer the possibility of circNDST1 as a stable serum marker for lymph node metastasis prediction in thyroid cancer. Mechanistically, we established through rigorous experiments that CSNK2A1, a kinase, acts as an important bridge between circNDST1 and the PI3K–Akt pathway. This also renders CSNK2A1 a feasible new biomarker and future therapeutic target for thyroid cancer, paving the way for novel drug interventions in thyroid cancer.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.